Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

3-Agent 6-Landmark

Benchmarks

Task NameDataset NameSOTA ResultTrend
3-Agent 6-Landmark3-Agent 6-Landmark Expert
Mean Episode Return683.3
12
3-Agent 6-Landmark3-Agent 6-Landmark Medium Expert
Mean Episode Return678.7
12
3-Agent 6-Landmark3-Agent 6-Landmark Medium
Mean Episode Return532.5
12
3-Agent 6-Landmark3-Agent 6-Landmark Medium Replay
Mean Episode Return542.4
12
3-Agent 6-Landmark3-Agent 6-Landmark Random Medium
Mean Episode Return329.9
12
3-Agent 6-Landmark3-Agent 6-Landmark Random
Mean Episode Return470
12
Multi-Agent Navigation3-Agent 6-Landmark Standard Environment (Random Medium)
Number of good policies13
8
Multi-Agent Navigation3-Agent 6-Landmark Standard Environment (Expert)
Number of Good Policies20
4
Multi-Agent Navigation3-Agent 6-Landmark Medium Expert Standard Environment
Number of Good Policies19.2
4
Multi-Agent Navigation3-Agent 6-Landmark Standard Environment Medium Replay
Number of Good Policies14.6
4
Multi-Agent Navigation3-Agent 6-Landmark Standard Environment (Random)
Number of Good Policies12.8
4
Showing 11 of 11 rows